Date of Award


Degree Type


Degree Name

Bachelor of Engineering Honours


School of Mathematics, Information Techology and Engineering


Faculty of Science, Technology and Engineering

First Advisor

Dr Zhihong Man


In this thesis, the problem of robust variable structure control for non-linear rigid robotic manipulators is investigated. Robustness and convergence results are presented for variable structure control systems of robotic manipulators with bounded unknown disturbances, nonlinearities, dynamical couplings and parameter uncertainties. The major outcomes of the work described in this thesis are summarised as given below. The basic variable structure theory is surveyed, and some basic ideas such as sliding mode designs, robustness analysis and control1er design methods for linear or non-linear systems are reviewed. Three recent variable structure control schemes for robotic manipulators are discussed and compared to highlight the research developments in this area. A decentralised variable structure model reference adaptive control scheme is proposed for a class of large scale systems. It is shown that, unlike previous decentralised variable structure control schemes, the local variable structure controller design in this scheme requires only three bounds of the subsystem matrices and dynamical interactions instead of the upper and the lower bounds of all unknown subsystem parameters. Using this scheme, not only asymptotic convergence of the output tracking error can be guaranteed, but also the controller design is greatly simplified. In order to eliminate chattering caused by the variable structure technique, local boundary layer controllers are presented. Furthermore, the scheme is applied to the tracking control of robotic manipulators with the result that strong robustness and asymptotic convergence of the output tracking error are obtained.

Included in

Robotics Commons